Activaton in Drugs

Drug Mechanochemical Activation
I. COLOMBO,1 G. GRASSI,2,3 M. GRASSI4

1
Eurand S.p.A., Physical Pharmacy Laboratory, via Martin Luther King, 13-20060 Pessano con, Bornago, Milano, Italy
Department of Internal Medicine, University Hospital of Trieste, Cattinara I-34149, Trieste, Italy
Department of Life Sciences, University of Trieste, Trieste, Italy
Department of Chemical, Environmental and Raw Materials Engineering DICAMP, University of Trieste,
Piazzale Europa 1, I-34127, Trieste, Italy
Received 30 September 2008; revised 12 December 2008; accepted 28 January 2009

Published online 31 March 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21733
ABSTRACT: The aim of this review is to describe the theoretical background lying
behind the solid drug mechanochemical activation by cogrinding pointing out its
advantages and drawbacks. A brief historical introduction precedes the discussion about
the mechanisms leading to solid drug activation. This allows to clarify the concept of
solid activation whose main effect is to improve drug solubility and, thus, drug bioavailability. Then, the attention is focused on the experimental tools used to evaluate drug
activation before the in vivo use. This, of course, permits to properly modulate the
milling conditions (milling time, mill revolution speed, drug/carrier ratio and so on) in
the light of the optimisation of milling process and activated system properties. Thereafter, the discussion shifts on the different kinds of mills that can be used and on mills
classification based on the energy transferred to the materials. Fundamental tool to
perform this task is the mathematical modelling of mill dynamics that is here shown for
different mills kinds. Finally, some examples of activated systems performance both
in vitro and in vivo are presented and discussed. In conclusion, mechanochemical
activation improves drug bioavailability. Interestingly, this activation does not require
the use of solvents whose elimination from the activated product can be difficult and
expensive but a relatively simple mechanical treatment. On the other hand, this
approach, usually, works only for poorly water soluble drugs (solubility <100 mg/mL)
that do not exhibit permeability problems. 2009 Wiley-Liss, Inc. and the American
Pharmacists Association J Pharm Sci 98:39613986, 2009
Keywords:
mechanical activation; cogrinding bioavailability
HISTORICAL BACKGROUND
Electrochemistry and mechanochemistry represent the two branches into which the nontherCorrespondence to: M. Grassi (Telephone: 39-040-5583435;
Fax: 39-040-569823; E-mail: mariog@dicamp.univ.trieste.it)
Journal of Pharmaceutical Sciences, Vol. 98, 39613986 (2009)
2009 Wiley-Liss, Inc. and the American Pharmacists Association
mally activated chemistry can be subdivided in. In

particular, mechanochemistry deals with the
physico-chemical transformations and chemical
reactions affecting substances following the
administration of mechanical energy.1 As mills
represent typical energy suppliers, grinding and
cogrinding are the common processes inducing
mechanochemical transformations.2 In virtue of
its intrinsic availability, mechanical energy has
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009
3961
3962
COLOMBO, GRASSI, AND GRASSI
been the first type of energy deliberately employed

by the human mind as witnessed by the fabrication of prehistoric weapons and fire production by
friction. Nevertheless, modern mechanochemistry
was born in between the end of the 19th century
and the beginning of the 20th century when the
first books on this topic were published and
the term mechanochemistry was, introduced in
the scientific literature.3 Undoubtedly, the
middle of the 20th century sees a considerable
development of solids mechanochemistry due to
the studies about explosion excitation in both
initiating and brisant explosive substances under
mechanical action. France, England and Russia
are the leading countries in this field and the
theory of hot spots, explaining the initiation and
development of explosion by local increase of
temperature at the contacts of two solid undergoing mechanical action, dates back to this period.
The development of new methods for mineral raw
processing, preparation of new construction
materials, mineral fertilizers and functional
ceramics, represented the goal of mechanochemistry researchers in Germany, Japan, Israel and
USSR in the 1960s.2 The end of the 1960s sees the
ingress of mechanochemistry in the material
science field with the ball milling production of
nickel- and iron based super-alloys, impossible to
be obtained by conventional melting and casting
techniques.4 Interestingly, the most important
mechanochemistry development, occurring in the
1980s,510 sees the end of the leading role played
by eastern block countries due to the renewed
interest of Japanese researchers and the rapid
progresses in mechanical alloying. The constitution of the International Mechanochemical
Association in 1988 and the 1st International
Conference on Mechanochemistry (Kos ice, Slovak
Republic, 1993), definitively consecrated Mechanochemistry all over the word. In the light of this
frame, it is not surprising that the pharmaceutical
interest for mechanochemistry developed only in
the 1980s.11 In addition, initially, the severe
requirements needed by pharmaceutical production (Good Manufacturing Practice and FDA
approval) hindered the clear affirmation of
mechanochemstry in this new field. Nevertheless,
once one of the most important problems (milled
material contamination by grinding media and
mill walls) was solved by adopting proper mill
lining and high abrasion resistant materials for
grinding media, the pharmaceutical doors were
definitively opened to mechanochemistry.12 Since
the beginning, this synergistic cooperation has
considerably developed and led to patents and

industrial applications.13,14 Basic reasons for this
fruitful cooperation rely on the possibility of
producing pharmaceutical products avoiding the
use of solvents (whose elimination can be difficult,
expensive and can alter drug activated status) and
the possibility of increasing the bioavailability of
poorly water soluble drugs.
MECHANICAL ACTIVATION
One Component Systems
A one component solid material being processed in
a mill receives mechanical energy in pulse form
every times it is trapped between two (or more)
colliding grinding media or between mill wall and
one (or more) wall-impacting grinding medium
(see Fig. 1). In particular, mechanical energy is
transferred by means of normal and shear
stresses acting on solid material surfaces. The
effect of this externally imposed stress field is the
growth of a strain field in the solid bulk. The strain
field manifests through different phenomena such
as (1) atoms shifts from equilibrium stable
positions at lattice nodes, (2) changes of bond
lengths and angles and, sometimes, excitation of
electron subsystem.2 Despite its apparent, macroscopic, simplicity, mechanical energy transfer into
solid material is very complicated. Among the
many theories developed to address this problem,
one of the most popular and interesting is the socalled triboplasma approach.1 Basically, it
assumes that an impact of sufficient intensity
results in a quasi-adiabatic local energy accumulation (temperature can grow up to 104 K in
submicroscopic zones of the impact). This, in turn,
gives origin to a metastable structure that must
release part of the accumulated energy to get a
more stable thermodynamic condition. According
to this approach, a multistage pattern of energy
dissipation takes place. Within the first 1011
107 s stochastic reactions dissipate the energy
pertaining to the triboplasma highly excited
energy states. After 107 s numerous elementary
excitation processes such as the recombination of
plasma products, the propagation and interaction
of dislocations, the propagation and emission of
electrons and photons and the formation of hot
spots during heat release occur. It is in this period
that phase transformations and mechanochemical
reactions proceed. Of course, part of the mechanical energy provided is retained by the solid
DOI 10.1002/jps
DRUG MECHANOCHEMICAL ACTIVATION
3963
Figure 1. Schematic representation of what happens inside mill bowl during grinding. Grinding media transfer mechanical energy to the charge (solid drug) in pulse form
as, in each collision, only a small fraction k of the total charge is involved.
material in the form of the so-called excess Gibbs

energy. Basically this energy is linked to longliving mestable states that are preserved in the
activated material after mechanical treatment.
This excess energy may be released very slowly
according to sequence of irreversible processes
approaching thermodynamic equilibrium. From
macroscopically point of view, energy relaxation
takes place according to three mechanisms: (a)
heat, (b) plastic deformation and (c) rupture of
chemical bonds (mechanochemical reaction).2
Due to the periodic impulsive action of grinding
media, the externally imposed stress field is
pulsating so that milling process sees an alternation of strain field formation and relaxation.
Obviously, the relative speed of the energy
relaxation processes and the stress field application plays an important role in determining the
final properties of the ground product.15
The main part of the supplied energy is
converted into heat. The concentration of the
strain field in particular crystal sites can lead to
crystal crushing and thus to the formation of new
DOI 10.1002/jps
surface. The iteration of this phenomenon induces

crystal size reduction (particle size reduction) to
some critical threshold. Further energy supply
yields to the accumulation of defects into crystal
volume or on its surface to finally lead to a
complete amorphisation. Although the molecular
solids amorphisation process remains unclear,16
in general, crystal transformation to amorphous
phase can be explained on the basis of two leading
theories: mechanical and thermodynamic destabilisation. According to the first theory,17 crystal
lattice collapse occurs for too high anharmonicity
of lattice vibrations (phonons) induced by compression. Indeed, in this condition, Born stability
criteria for crystal lattice are violated. The
second theory18 affirms that mechanical energy
continuously increases crystal defects concentration up to a critical threshold beyond which
amorphous phase is thermodynamically more
stable than the crystal one. Anyway, amorphisation process implies the formation of defects such
as point defects (e.g. interstitial occupancies and
vacancies), line defects (e.g. edge and screw
3964
dislocations), plane defects (e.g. grain boundaries

and crystal surfaces), and anomalous distribution
of punctual defects1921 that can decrease material true density due to the formation of the
so-called activation volume.20 Experimental
evidences22 indicate that crystal amorphisation
starts on a thin surface layer (0.11 mm20) and
then propagates into the bulk. Undoubtedly, this
internal propagation can be favoured by intrinsic
crystal disorder (due to molecules crystallising in
a single conformation but in a different orientation relative to other molecules in the crystal)
typical of pharmaceutical solids such as nonsteroidal antiiflammatory drugs, salicylsalicylic acid
and antiarrhythmic compound flecainide.23
If defects formation is not random but it follows
particular orders, a metastable polymorph structure occurs.5 When the preferred relaxation way is
the third one (rupture of chemical bonds), a
mechanochemical reaction occurs.10 The term
mechanochemical activation means the accumulation of defects (amorphisation process), the
formation of polymorphs and chemical reactions
occurrence. Accordingly, the intrinsic energy
content of an activated system is higher than
that of a not-activated one. Thus, the reason for
the adjective activation relies on the potential
ability of the system of doing something. Indeed,
transformations of a high energy content system
are facilitated as the energetic barriers hindering
the evolution to a new equilibrium condition are
lower than those of a not-activated system which
is thermodynamically more stable. Accordingly,
the term activation expresses an intrinsic potentiality of the activated system.
Comminution
In the light of what above discussed about
mechanochemistry, the grinding of a solid (crystalline) material leads to: (1) particle size reduction (plastic deformation leading to comminution),
(2) mechanical activation and (3) heat release.
Whereas the last two aspects have been previously discussed, a basic interpretationprediction of comminution can be easily given according
to the Griffith theory.7,24 The rupture of an ideally
brittle material implies the interruption of interatomic bonds with the consequent formation of two
new surfaces. Obviously, the breaking energy Be
per unit surface (as2/E, where a is the distance
between two consecutive crystalline planes, s is
the applied stress and E is Young modulus) must
be equal to the surface energy, 2gsv, competing to

the two newly formed surfaces. Accordingly, the
critical stress sc leading to rupture is:
r
2Eg sv
(1)
sc
a
Griffith explained the huge equation (1) overestimation of experimental sc values affirming
that each material contains many elliptical cracks
reducing mechanical resistance. Indeed, he postulated that at the tip of such cracks (corresponding to the end of the major axis) a strong
concentration of the stress occurs. On this basis,
he found that the energy per unit length, Le,
released due to crack propagation is:
pc2 s 2
plane stress;
E
1 m2 2 2
Le
pc s plane strain
E
Le
(2)
where m is the Poisson modulus (0.5 for incompressible materials), c is the half length of the
crack major axis and s is the stress applied to the
bulk. Obviously, crack propagation (enlargement)
implies the formation of new surfaces and,
consequently, an increase of system surface
energy Se 4cgsv. According to Griffith, the crack
will propagate, producing brittle fracture, if crack
propagation does not imply system energy
(Se Le) increase (d(Se Le)/dc 0). This leads
to the following Eq. (1) improvement:
r
2Eg sv
plane stress;
se
pc
s
(3)
2Eg sv
plane strain
se
p1 m2 c
As real crystals do not show ideally brittle fracture
(crack propagation is accompanied by material
plasticisation near the crack tip), Eq. (3) still
detaches to real behaviour. Nevertheless, if gsv is
replaced by the effective surface energy gsve, sum
of gsv and gpl (local plastic deformation energy7),
Eq. (3) yields satisfactory agreement with experimental data. gsve depends on internal and external
factors. Among the external factors, magnitude
and mode of stress application (normal or shear),
rate of strain and temperature must be mentioned. For what concerns internal factors,
material structure, impurity concentration, grain
and particle size and the extent of preceding
plastic strain must be remembered.7 It is
worth mentioning that for brittle materials, gsve
DOI 10.1002/jps
coincides with the material solidvapour surface

tension whose determination is, usually, not so
cumbersome.25 The opposite occurs for the tough
material (gsvegpl).
The importance of comminution does not
exhaust in enhancing drug dissolution kinetics
due to the increase of solid surface per unit area.
Indeed, it is well known25 that solid solubility
depends also on crystals size and the reference
equation for this effect is the OstwaldFreundlich
relation:26,27

RT
Sr
2g
ln
(4)
r
sl
M
S1
r
where r and M are, respectively, solid density and
molecular weight, R is the universal gas constant,
T is the absolute temperature, Sr and S1 are,
respectively, the solubility of a spherical crystal
characterised by radius r and infinite radius (i.e. a
plane) and gsl is the solidliquid surface tension. It
is now important to spend few words on the
concept of crystal radius. Ground material
appears as a distribution of different radius
particles (secondary grains) each one constituted
by an ensemble of crystals (primary grains) bound
together by amorphous connecting phase. Indeed,
mechanical action produces an increase in lattice
defect density leading to the formation of coherent
crystalline domains (crystallites) inside each
crystal. Accordingly, in Eq. (4), r refers to the
crystallite radius. Of course, the frame is more
complicated by the fact that real crystallites are
not necessary spherical and, in this case, r should
refer to crystallite local curvature radius.28
Although Eq. (4) has been subject to severe
criticism,29,30 experimental evidences show that
solid solubility increases for sufficiently small
crystallites (usually r < 20 nm). For example, it
has been reported that Griseofulvin solubility in
water (T 378C) increases from 11.9 mg/cm3
(macrocrystal) to 60.2 mg/cm3 for nanocrystals,25
methylhydroxyprogesterone acetate (MPA) solubility in water (T 378C, pH 5.5) increases from
1.2 mg/cm3 (macrocrystal) to around 3.5 mg/cm3 31
and nimesulide solubility in water (T 378C;
pH 5.5) increases from 11 mg/cm3 to around
25 mg/cm3.32 Accordingly, we believe in the
qualitative correctness of Eq. (4) even if nothing
can be said on its quantitative reliability in the
light of the high experimental difficulties involved
in its verification. Indeed, whenever a distribution
of crystal size exists, Ostwald ripening, caused by
the varying curvature of different interfaces, can
lead to the dissolution of less stable smaller
DOI 10.1002/jps
3965
crystals that favour the growth of larger crystals.33 Nevertheless, the increased solubility of
very small crystals induces to enlarge the
concept of mechanical activated systems also to
nanocrystals.
Multicomponents Systems
In One Component Systems Section we focussed
the attention on the mechanisms leading to
mechanochemical activation in the case of one
component grinding. In particular, we saw that
the mechanical energy supplied makes the
material metastable so that it relaxes to a more
stable condition by releasing heat, breaking
(particle size reduction) and giving origin to an
activated material. Typically, these relaxation
phenomena are very rapid7 (101107 s) and this
is the reason why the activated material is
considered stable. Unfortunately, however, thermodynamically speaking, nanocrystals, polymorphs, amorphous and defects containing
materials are, usually, far from being stable12 at
least with respect to the lifetime required to a
pharmaceutical product (months or years).
Accordingly, the addition of at least one more
component (stabilizer) to the grinding mixture is
required. Indeed, mechanically treated drugs can
interact, for example, via Van der Waals or
hydrogen bonds, with the stabiliser yielding to
pharmaceutically stable products conserving
their activation.12 Obviously, the presence of the
stabilizer makes the grinding scenario more
complex as mechanical energy exerts its action
on the drug and on the stabilizer inducing
transformations in both of them. For example, if
the stabiliser is represented by a crosslinked
polymer, drug activation and interaction with a
changing particle size stabiliser simultaneously
occur.32 Example of drugstabiliser couples are
diethylstilbestrol, indomethacin, griseofulvin,
resorcin, ibuprofen, with b-cyclodextrin, polyethylene glycol, polyvinylpyrrolidone and methylcarboxycellulose.16 Figure 2 shows a schematic
representation of the possible spatial relations
between a drug (amorphous drug, spheres; macroor nano-crystalline drug, white structures) and a
cyclodextrin (grey cuplike structures). While
amorphous drug can be found inside cyclodextrin
cavity (inclusion complex) or among different
cyclodextrin units, bigger drug crystals can be
found only among different units. The stabilising
action is exploited through physical interactions
between drug and cyclodextrin surfaces. If the aim
3966
Figure 2. When cyclodextrin is used as carrier in

cogrinding, its stabilising action manifests through
surface interactions with amorphous or macro
nano-crystalline drug or through the formation of inclusion complexes.
of the delivery system is a prompt action, the

inclusion complex situation is not to be preferred
as, reasonably, drug release will be slower with
respect to the external drugcyclodextrin interaction situation. Figure 3 reports the microscopic
aspect of a coground system composed by
the polymeric stabilizer (micronised crosslinked
Figure 3. SEM picture of a coground mixture composed by a drug (needles) and a crosslinked polymer
(particles). Drug needles can stand alone on polymeric
particles surface or inserted in the polymeric network.
Figure insert shows the EDS spectrum referring to the
drug (thick line, point (N), drug needles) and to polymeric particle where drug needles cannot be detected
(thin line, point (M)). As both the spectra are qualitatively similar, the drug presence inside the polymeric
matrix (M) is confirmed.
polyvinylpyrrolidone particles) and an antiinflammatory drug constituted by oxygen, fluorine, carbon and sulphur. Interestingly, an energy
dispersive spectroscopy (EDS) analysis, permitting to identify the EDS compound spectrum
(determined by its constituent atoms), allows to
verify that the drug is present both outside and
inside polymeric matrix (globular structures in
Fig. 3). Indeed, EDS spectrum reveals that
sulphur peak (an element that the drug and the
polymer do not share) is present in both needle
structures (point N, thick line, Fig. 3) and in
polymeric matrix (point M, thin line, Fig. 3)
where no needles (drug crystals) are visible.
Accordingly, not only drug crystals are stabilised
by surface interactions with the polymer, but part
of the drug, lying inside particles, is stabilised by
the polymeric network hindering drug transformations. Finally, it is worth mentioning that as
drug crystals (needles) depicted in Figure 3 are
almost cylindrical shaped, they can be considered
nanocrystals in correspondence of the cylinder
basis where the surface curvature radius is lower
than 300 nm.
EXPERIMENTAL VERIFICATION
OF ACTIVATION
The problem of the experimental verification of
mechanochemical activation falls in the broad
field of solid-state characterisation. Accordingly,
techniques relying on differences in periodicites of
atoms in crystals (X-ray powder diffraction
XRPD), energies of bond stretching/bending
vibrations and lattice vibration (IR, Raman),
electronic environments of nuclei (nucleic magnetic resonance, NMR), heat flow or weight
change (thermal analysis: differential scanning
calorimetry (DSC) and thermal gravimetric analysis (TGA)) and morphology (optical microscopy),
can be useful at this purpose.34,35 In particular, in
the mechanochemical activation context, solid
state characterisation serves to exclude the
formation of new chemical entities (occurrence
of chemical reactions) and drug polymorphs. At
the same time, this characterisation is needed to
estimate the residual amount of drug crystallinity
(Xrc) after mechanical treatment. Indeed, this
parameter can be roughly elected as a measure of
drug activation. To be more precise, a deeper
evaluation of activation not only requires the
determination of Xrc but also the size distribution
of drug crystals (it is worth remembering that
DOI 10.1002/jps
drug solubility increases for smaller crystals, see

Eq. 4). In addition to these techniques, other
approaches can provide estimation of mechanochemical activation. Among others, solution
calorimetry, water sorption36 and release tests
(RT) can be mentioned.25 As the description of all
the above mentioned approaches for the evaluation of mechanochemical activation is out of the
aim of this review, we focus the attention on those
commonly used in the industrial field.
DSC
Some nano-sized crystals (or crystallites) peculiar
behaviour depend on the fact that their properties
are influenced by their surface atoms rather than
by their bulk atoms. Indeed, it is well known that
surface atoms have fewer interatomic bonds than
those in the bulk, this making them more loosely
bound than bulk atoms.37 Huang et al.38 revealed,
by experimental data and molecular dynamics
simulations, neat differences in the structural
dynamics of nanocrystals surface and bulk atoms.
In particular they found that coherent diffraction
patterns recorded from individual nanocrystals
are very sensitive to the atomic structure of
nanocrystal surfaces. Assuming, for the sake of
simplicity, a spherical shaped crystal, all we know
that the surface/volume ratio becomes infinite for
vanishing crystal radius and this is the reason
why the importance of surface atoms becomes
more and more pronounced as crystal radius
decreases. Typically, nano-crystal melting temperature (Tmr), melting enthalpy Dhm, solid
vapour (gsv) surface tension and grain growth
rate (GGR) depend on nano-crystal radius and
differ from those of the infinite crystal (macrocrystal). Nano-crystals melting temperature
deviations are related to interfaces nature3941
and curvature.40 The melting point shifting in
nano-sized materials was originally predicted by
the thermodynamic nucleation theory in the early
1900s (GibbsThomson equation).27,42 More
recently, Brun et al.,43 starting from the Laplace
and GibbsDuhem equations, found the following
relation between melting temperature and crystallite radius:
Z1=r
Dhmr
dT 2
T
0
Tm1

1
1
1
g
1
g lv

d
sl d
rs rl
Rlv
Rsl
rs
ZTmr
DOI 10.1002/jps
3967
where Tm1 is the infinite radius crystal melting

temperature, glv and gsl are, respectively, liquid
vapour and solidliquid surface tensions, rs and rl
are, respectively, solid and liquid drug density, Rlv
and Rsl are, respectively, liquidvapour and solid
liquid interfaces curvature radii while r is crystallites radius. If drug crystallites are much more
abundant than amorphous boundaries (then they
are tightly packed), it can be assumed that
melting starts with each crystallite covered by a
liquid skin. Accordingly, we have that Rlv Rsl r
(crystallite radius). On the contrary, if the
amorphous drug fraction is much larger than
the crystalline one, crystallites melting takes
place in an amorphous (liquid) bath (we
remember here that amorphous phase becomes
liquid as soon as its glass transition temperature
is exceeded and this happens much before crystallite melting), and, thus, it can be assumed Rlv 1
and Rsl r. On the basis of a thermodynamic cycle
interpreting crystallite melting process as the
sum of five steps ((i) crystallites clustering at Tmr
to form the macro-crystal; (ii) macro-crystal
heating from Tmr to Tm1; (iii) macro-crystal
melting at Tm1; (iv) liquid disintegration into
nano-sized droplets at Tm1; (v) droplets cooling
from at Tm1 to Tmr) Zhang et al.37 demonstrated
that Dhmr dependence on crystallite radius is
given by:
Dhmr

TZm1
3 g sv g lv
Dhm1

DCp dT (6)

r rs
rl
Tm
where Dhm1 is macrocrystal melting enthalpy

while DCp is the difference between the liquid and
the solid specific heat capacity at constant
pressure. Eqs. (5) and (6) simultaneous numerical
solutions are shown in Figure 4 in the case of
nimesulide (Tm1 148.78C, Dhm1 108,720
J kg1, rs 1490 kg m3, rl 1343.7 kg m3,
glv 44.3 103 J m2, gsl 13.3 103 J m2,
Rlv 1 and Rsl r25). The fact that they predict
a reduction of both Tmr and Dhmr with decreasing r
can be physically interpreted as the increased
facility of network braking due to the presence of
surface atoms that are more loosely bound than
bulk atoms.37 It is evident that for r 10 nm, a
significant melting temperature and enthalpy
reduction occur. Remembering that nimesulide
unit cell side is approximately 0.87 nm,25 Eqs. (5)
and (6) suggest that appreciable melting temperature reductions take place when mean
crystallite radius (calculated on the volumetric
3968
Figure 4. Theoretical dependence of nimesulide

melting temperature (Tmr black line) and melting
enthalpy (DHmr grey line) on crystals radius r.
distribution vs. radius) is constituted by approximately 20 unit cells. Figure 5 shows the experimental DSC pattern referring to pure nimesulide
and to a nimesulide/crosslinked polyvinylpyrrolidone mixture (1:3, w/w) after 1 h grinding in a
planetary mill. Melting temperature reduction
witnesses the absence of the original macrocrystals (no thermal event is now detectable at
148.78C) and the presence of an ensemble of nanocrystals whose mean melting temperature is
126.68C. It is worth mentioning that DSC can

be also very important for a full characterisation
of drug status after cogrinding. Indeed, it
allows the estimation of the fraction of original
macro-crystals, nano-crystals and amorphous
phase. This can be performed comparing the
specific melting enthalpy of an un-ground drug/
carrier mixture and the specific melting enthalpy
of the same mixture after cogrinding. In addition,
DSC allows also the determination of nanocrystals size distribution.
Magomedov,44 studying the surface energy
dependence on nano-crystal size and shape, found
that surface energy decreases with crystallite
radius and this behaviour becomes more and more
important for smaller nano-crystals. Interestingly, he noticed that nanocrystal melts when
the surface energy decreases to a value at which it
becomes independent of the nanocrystal size and
shape.
Another peculiar aspect of nano-crystals is the
GGR. Driving force for nucleation and growth of
crystallites is the free energy difference between
amorphous and crystalline phases and this
difference mainly depends on temperature and
crystallites diameter. For its practical and theoretical importance, researchers have deeply studied this problem in order to derive equations able
to predict crystallites rate growth. Although more
refined approaches can be considered, GGR can be
retained proportional to the inverse of the crystal
diameter (Dc) powered to an exponent n greater
than 1:45
dDc kT
n ;
Dc
dt
Dc
Dn1
c0
(7)
n 1kTt
1=n1
where Dc0 is initial crystallite diameter. Another

way of matching the problem of crystallites
growth is considering the JohnsonMehlAvrami
equation (JMA), or its modified form,46 providing
the increase of crystal fraction x versus time t:
x 1 expkt tn
Figure 5. Differential scanning calorimeter pattern

(DSC) referring to original nimesulide (melting temperature 148.78C) and nimesulide coground with crosslinked polyvinylpyrrolidone in a planetary mill for 1 h in
a ratio 1:3 (w/w) (melting temperature 126.68C).
(8)
where k is a temperature dependent rate constant,

t is an empirical parameter named scaling factor
and n is another empirical factor. The fact that Eq.
(8) can also predict a very slow initial increase of x,
is motivated by the fact that strong hydrogen
bonding difference between amorphous and
crystalline states can delay crystallisation and
vice versa.47
DOI 10.1002/jps
XRPD
When, upon melting, material decomposes and/or
solid phase interactions occur during heating,
DSC cannot be applied to check the presence and
abundance of crystals. In this case, a valid
alternative method is represented by XRPD
provided that the amount of amorphous phases
and/or adsorbed water is not excessive and
provided that the X-rays diffraction pattern is
not affected by severe peak overlaps and preferred
orientation of the powder sample is not significant. Although a detailed description of crystals
abundance determination after drug/carrier
grinding goes behind the scope of this review, it
is interesting recalling that the working equation
is:48
P
IiD
XD P i
f XD a
(9)
I
iD
i
0
where XD is drug crystalline fraction in the ground
mixture, IiD is the area of the ith peak belonging
to
P
the
ground
drug
X-rays
pattern,
I
and
i iD
P
i IiD 0 are, respectively, the sum of all the peak
areas characterising the ground and native drug
X-rays pattern, a is a semi-empirical factor taking
into account micro absorption effects and f(XD) is a
function of XD depending
P on the drugcarrier
mixture. The unknowns i IiD 0 , a and f(XD) can
be calculated from a calibration procedure based
on the measurement of several physical mixtures
of the native crystalline drug and carrier.
It is important to remember that if amorphous
structures produce broad diffraction reflections
owing to the lack of any long range periodicity in
the atomic arrangement, nanocrystalline materials comport a deviation from the X-ray diffraction
pattern competing to an ideal crystal as atomic
arrangement periodicity exists only on few
molecular distances. In particular, finite crystallite size, strain and extended defects (stacking
faults and dislocations) lead to broadening of the
diffraction peaks. In this contest, whole pattern
profile modelling (peaks positions, intensities,
width and shapes) can provide a complete
evaluation of grain size distribution and lattice
defect content.25 The use of the so-called radial
distribution function, determinable from the
experimental patterns by means of the Fourier
transform, allows determining the probability of
finding an atom at a given distance from the
centre of a reference atom of the system.49 Very
smooth and low noise experimental data are
needed to achieve reliable and reproducible
DOI 10.1002/jps
3969
results. For these reasons synchrotron and

similarly intense X-ray sources are recommended.
An alternative approach to the analysis of
diffraction data of amorphous materials has been
recently proposed.50 In this framework, the
crystallite size, jointly to microstrain, has been
introduced to account for the broadening of the
diffraction peaks of an amorphous structure. By
reducing the coherently scattering domains, the
corresponding Bragg reflections will broaden to
such an extent to give place to a typical amorphous
diffraction pattern (halo).
Finally, we would like to conclude this section
reminding an important function of XRPD in the
field of mechanical activated systems. Discussing
the meaning of activated systems (see One
Component Systems Section), we said that
mechanical energy supply can lead to the occurrence of chemical reactions. Whereas often this is
the aim of the mechanical treatment, in the
pharmaceutical field this is not desirable. What is
needed is drug activation without the formation of
new chemical entities or polymorphs that would
lead to huge regulatory problems. Indeed, FDA
approval for the native drug is absolutely not
extensible to other chemical compounds generated by the mechanochemical treatment. In this
frame, XRPD plays a very important role as the
formation of drug polymorphs or the occurrence of
drug chemical transformations upon grinding
reflect in evident differences between the XPRD
pattern of native and ground drug.
Release
Drug release test from activated systems is
another excellent method to evaluate the activation grade even if the determination of the
crystalline, nano-crystalline and amorphous drug
abundance is not direct and can be indirectly
determined interpreting experimental data by
proper mathematical models.25 It does not exist a
unifying model describing drug release from every
activated systems (but all of them must share
some common characteristics) as release kinetics
strongly depends also on the carrier (stabiliser)
considered. In this light, carriers can be roughly
subdivided into two main classes: crosslinked
polymers and the rest of the world. This subdivision is motivated by the fact that while in the
second case drugcarrier interactions develop
superficially through more or less complex
adsorption/desorption mechanisms, in the first
3970
case (crosslinked polymers) also topological carrier properties can play a very important role as
the drug can be embedded inside the (stable)
polymeric network. Accordingly, both drug diffusion in the (swelling) network and drug-polymer
interactions affect release kinetics. Beside these
mechanisms, also the crystalline, nano-crystalline
and amorphous drug dissolution into release
medium heavily rules release kinetics. Indeed,
upon contact with the external release medium,
the stabilising action exerted by the carrier
vanishes (obviously, with different kinetics
depending on carrier type) and the drug starts
dissolving with the possibility of macro-crystals
and/or polymorphs formation. As drug solubility
depends on crystal radius (amorphous phase can
be regarded as a crystal of vanishing radius, see
Comminution Section), the release process takes
place as the drug were characterised by a time
dependent solubility.25 One of the most commonly
used equations to describe this particular behaviour is:51
Cs t Ca;nc Cmc eKr t Cmc
(10)
where Cs is the time (t) dependent drug solubility,

Kr is the re-crystallisation constant, Ca,nc represents the amorphous or nanocrytsalline solubility
while Cmc indicates the solubility of macrocrystals. This equation simply states that at the
beginning (t 0) Cs is equal to Ca,nc and then it
exponentially decreases to get Cmc. Obviously, Eq.
(10) intrinsically accounts also for macro-crystals
dissolution (those who do not undergo re-crystallisation phenomena). Indeed, their dissolution is
represented by Eq. (10) setting Kr 1 (Cs is
always equal to Cmc). Accordingly, the general
dissolution equation becomes:
@Ca;nc
Kta;nc Cs t C
@t
conditions according to usual dissolution testing.

If in sink conditions the higher activation grade
reflects into higher release kinetics, the adoption
of un-sink allows seeing a peculiar release
pattern. Indeed, due to the conversion of amorphous drug into the more stable macro-crystalline
form, implying drug solubility reduction, the
release concentration curve can show a rapid
increase followed by a decrease as shown in
Figure 6 in the case of an anti-inflammatory drug
(Cmc 2.5 mg/cm3; amorphous drug fraction 0.6,
nanocrystalline drug fraction 0.4) release from
coground drug/crosslinked polyvinylpyrrolidone
(1:2, w/w) system under un-sink conditions.
Stability
An essential prerequisite for the pharmaceutical
use of mechanochemical activated systems is their
physical stability over months or years. At this
purpose, specific stability tests are performed in
temperature and humidity controlled environments. Typically, DSC, XRPD and release tests
are conducted just after sample preparation and
at fixed times after exposure to constant temperature and humidity environments. Temperature can span from 5 to 408C while humidity can
be up to 90%. If activation characteristics such as
(11)
where Kta;nc indicates the amorphous or nanocrytsalline drug dissolution constant. Of course, three
Eq. (11) dissolution types exist: the first referring
to the amorphous drug, the second referring to
nano-crystals and the third referring to macrocrystals. The contribution of each equation to the
global dissolution flux depends on amorphous,
macro and nano-crystals relative abundance.
Coupling Eq. (11) with a proper equation accounting for drugcarrier interaction and, eventually,
drug diffusion through the polymeric network,
yields the final mathematical model formulation.
From an experimental point of view, release
tests can be performed in sink or un-sink
Figure 6. Amorphous drug re-crystallisation can

yield, in nonsink conditions, to the appearance of a
maximum in the release curve (release environment
concentration C vs. time t). This picture refers to the
release of a poor soluble drug (solubility Cs 2.5 mg/cm3,
water, 378C) coground with crosslinked polyvinylpirrolidone (1:2 (w/w) ratio). Amorphous drug fraction 0.6;
nanocrystalline drug fraction 0.4.
DOI 10.1002/jps
amorphous, nano-crystalline drug contents and

release kinetics are substantially stable over time,
the carrier (stabiliser) is approved. Obviously, in
particular cases, essentially dependent on drug
and carrier chemical characteristics, also other
chemical tests can be performed. The possibility of
adding other substances enhancing drug stabilisation can be also considered.
BIOAVAILABILITY ENHANCEMENTS
Bioavailability, defined as the rate and extent to
which the active drug is absorbed from a
pharmaceutical form and becomes available at
the site of drug action,52 depends on several
factors, among which drug solubility in an
aqueous environment and drug permeability
through lipophilic membranes play the role of
key parameters.53 Indeed, only solubilised molecules can be absorbed by the cellular membranes
subsequently reaching the site of drug action
(vascular system for instance). According to the
high or low values assumed by these parameters
(solubility and permeability), drugs can be classified into four different classes54 and a drug can be
defined bioavailable if it belongs to the fourth class
(high solubility and permeability). Many different
techniques are commonly used to improve the
bioavailability of poorly water-soluble but permeable drugs.55 In this sense, drug particle size
reduction, drug conglobation inside the lipidic
matrix of nano- or microspheres56 or drug
solubilisation in the dispersed lipophilic phase
of an O/W emulsion or microemulsion5760 can be
mentioned. A similar task can be achieved
complexing the drug with cyclodextrins by mixing
in solution or in presence of the melted drug.61
Finally, by means of solvent swelling it is possible
to load a drug into a polymeric carrier in a
nanocrystalline or amorphous state, thus considerably increasing its bioavailability.32,62 For
example, the simple particle size reduction (by
means of a milling process) allowed reducing
fenofibrate dose (Tricor1) from a 300 mg capsule
(standard drug) to a bioequivalent 145 mg tablet
containing nano-particulate drug.55 Interesting
examples of lipid based formulations (either selfemulsifying or emulsifying due to the presence of
bile salts) regards antivirals (Norvir1 (ritonavir)
and Fortonase1 (sanquinavir)) and immune
suppressant cyclosporine (Sandimmune1 and
Sandimmune Neoral1). If Fortonase1 increased
sanquinavir bioavailability up to threefold with
DOI 10.1002/jps
3971
respect to the original Invirase1 (sanquinavir

mesylate in powder form),63 the reduction of
emulsion particle size allowed Neoral1 to be more
bioavailable than the original Sandimmune1.64
Cyclodextrins can be found in several marketed
products such as Vfend1 (voriconazole), Geodon1
(zispradisone mesylate) and Sporanox1 (itraconazole). These solutions are intended for injection
or oral use55 and all of them are characterised by
high cyclodextrin/drug ratio (from around 15:1 to
40:1). Prograf1 and Sporanox1 capsules are
successful examples of commercial application of
the solvent swelling technique.65,66 Obviously,
each approach shows advantages and drawbacks
and it is more suitable for a determined drug or
for a specified administration route. Mechanochemical activated systems, in particular, can be
administered in the form of tablets or capsules, as
both formulations do not modify the activated
status. In general, any formulation that does not
require the use of solvents or high temperature
can in principle be considered. For these reasons,
mechanochemical activated systems are suitable
for oral administration, the most common route
for drug delivery into the human body because it
leads to a better patient compliance and is very
versatile for what concerns dosing conditions.
Nevertheless, mechanochemical activation and, of
course, solvent swelling activation, comport a
considerable bioavailability improvement if drug
solubility in aqueous medium is approximately
lower than 100 mg/cm3.
MILLS
Although it is often believed that a good mill is also
a good mechanical activator, this is not always
true as these two devices are intended for different
purposes. Indeed, if a mill is aimed to maximise
ground material specific surface (particle size
reduction) and to realise a good mixing with the
minimum energy expenditure in the shortest time
possible, mechanical activator target is to induce
defects (plastic deformation) in the ground material structure (see Comminution Section). Accordingly, an optimal mechanical treatment (OMT)
should imply an initial reduction of particle size
(milling) followed by the mechanical activation.
This is the reason why OMT requires the
sequential use of two different machines or the
same machine working with different operating
conditions.15 In order to meet all these requirements, many different grinding devices exist and
3972
they can be subdivided into three main categories

on the basis of how energy is transferred to the
material to be ground (mill charge):2,15 (1) ball
mills, (2) shear action mills and (3) shock action
mills. Although an exhaustive and complete
description of all the existing mills goes beyond
the scope of this review, some of the most relevant
examples will be discussed in the following
section.
Ball Mills
In ball mills, energy is transferred to grinding
media and mill charge (the mixture to be ground)
by mill body or by impellers. Mechanical action
takes place through both shear and normal
stresses and their relative importance can be
modulated in a wide range, acting on mill building
features and operating regime.67 Tumbling ball
mills, Planetary, vibrational, Spex mills and
attritors belong to this category.
Tumbling ball mills, relatively cheap, reliable
and easy to control and maintain, are constituted
by a rotating cylinder (drum), characterised by a
high length/diameter ratio, containing charge and
grinding media (balls). As charge grinding is due
to balls movement provoked by drum rotation, the
bulk energy supplied depends on drum diameter
and speed. They are primarily used for large-scale
industrial applications (for example cement
industry).
Planetary mill, deriving its name from the
planet-like movement of its vials, is essentially
made up by a circular basement rotating around
its main symmetry axis and carrying two or more
(even number) vials. These vials, containing
grinding media (balls), can be co- or counterrotating with respect to the basement. This
assembly induces grinding balls to run down
the inside wall of the vial (friction phase) to finally
collide against the opposing inside wall (impact
phase). Energy transfer to charge takes place both
in the friction and impact phase. Laboratory and
pilot-plant planetary mills can perform accelerations up to 100 times gravitational acceleration.
These mills are suitable for both ultrafine grinding and mechanical activation.7 Grinding vials
and balls are available in agate, silicon nitride,
sintered corundum, zirconia, chrome steel, CrNi
steel, tungsten carbide, and plastic polyamide.68
A typical high-energy vibrational mill is made
up of a bowl having, approximately, a toroidal
shape and containing the grinding media. The
upper part of the bowl is cut and covered by the
mill lid, while the lower part is connected by

springs to get a resiliently supported rigid body to
a metallic base fixed to the ground. Alternatively,
the bowl can be replaced by a metallic disc
carrying separate cylindrical grinding vials. The
bowl or the metallic disc are rigidly connected to
an (electrical) engine (placed approximately on
the bowl symmetry axis) determining the motion
of a shaft carrying, in its upper and lower parts,
two eccentrics. When the engine runs, the rotation
of the shaft provokes the motion of the two
eccentrics, which, in turn, give origin to a torque
determining the complex dynamics of the
bowl (metallic disc)/engine system. In the bowl
case, grinding media vibrate and undertake
a spiral movement around the toroidal bowl
axis.69 Grinding media acceleration essentially
depends on vibration motion frequency and
amplitude. Typically, frequency spans from 25
revolution/s down to 16 revolution/s while amplitude ranges between 2 and 12 mm. In these
conditions, grinding media acceleration does not
exceed 20 times gravitational acceleration. Beside
acceleration, grinding effect depends on grinding
media shape (typically balls or cylinders), density
and grinding media/charge ratio. Vibrational
mills, operating batch-wise or continuously,
are suitable for both grinding and mechanical
activation in pilot-plant and industrial applications.7 Mill bowl can be constituted, for example,
by stainless steel (with or without an internal
lining usually made up by rubber or polyurethane)
or polyurethane while grinding media are
typically made up by high abrasion resistant
material such as alumina (Al2O3), zirconia (ZrO2)
and agate.
SPEX mill, essentially used for laboratory
purposes, consists of one vial, containing the
charge and grinding balls, secured in the clamp
and swung energetically back and forth several
thousand times a minute. The back-and-forth
shaking motion is combined with lateral movements of the ends of the vial, so that the vial
movement resembles an 8 or 1. For each swing
of the vial, the balls clash against the sample and
the end of the vial. Because of the amplitude
(about 5 cm) and speed (about 1200 rpm) of the
clamp motion, ball velocities are on the order of
5 m/s and, consequently, the energy involved in
each impact is very high. Therefore, these are
high-energy mills. Hardened steel, alumina,
tungsten carbide, zirconia, stainless steel, silicon
nitride, agate, plastic, and methacrylate are used
for vials building and balls.68
DOI 10.1002/jps
Attritors are made up by a vertical or horizontal

drum carrying a series of impellers inside it and
positioned at right angles to each other. A proper
engine induces impellers rotation that is transferred to grinding media and charge. Mechanical
energy supply occurs because of impact between
balls, balls and drum wall and balls and impellers.
Nevertheless also charge interparticle collisions
and balls sliding can be responsible for mechanical
energy supply. Stainless steel or stainless steel
coated inside with alumina, silicon carbide, silicon
nitride, zirconia, rubber, and polyurethane constitute drum. Grinding media can be made up by
glass, flint stones, steatite ceramic, mullite, silicon
carbide, silicon nitride, sialon, alumina, zirconium
silicate, zirconia, stainless steel, carbon steel,
chrome steel and tungsten carbide. Laboratory
and industrial attritors can be found.68
Shear Action Mills

In shear action mills energy is given to crushing
elements (solid surfaces in relative motion) among
which mill charge lies. Rollers mill is a typical
example of shear action mill where mill charge is
forced to pass through two parallel counterrotating cylinders.
Shock Action Mills

In shock action mills energy is directly given to
mill charge. Jet and high peripheral-speed pin
mills belong to this category. In jet mills, no
moving parts exist as charge particles collisions
are due to a gas jet (compressed air or superheated
steam). Indeed, the charge is carried in a gas
stream flowing at high velocity where single
charge particles, colliding at approximately
1001000 m s1, undergo mutual attrition and
collisions. Both vertical and horizontal grinding
chambers may exist and the nozzles introducing
carrier and charge are located in different
positions. Sometimes (coarser grinding) they can
be positioned on opposite sides in order to
originate countercurrent streams. If comminution
mainly takes place near the nozzles where
particles collisions are more probable and energetic, the effect of particle collisions with chamber
wall and rigid reflection bodies cannot be
neglected. The most important advantages of jet
mills lie in their reduced dimensions easy of
maintenance and the possibility of coating the
grinding chamber with different liners. Typically,
DOI 10.1002/jps
3973
liners are constituted by polyurethane or dense

ceramic.
High peripheral-speed pin mills are constituted
by two counter-rotating rotors carrying, in concentric rows of circles, pin-breakers. Charge is
centrally fed so that, in their motion from centre to
periphery of rotors, particles collide with pin
breakers and with each other. Relative circumferential velocity can be up to 200300 m s1.
Number of concentric rows of circles, distance
among pins and their geometry, are key factors for
the product final characteristics.7
Cryogenic Mills
Cryomilling consists in milling materials at
cryogenic temperatures and/or it consists in
milling in presence of cryogenic media such as
liquid nitrogen (wet grinding).68 Despite the
considerable costs connected to the use of cryogenic fluids (e.g. nitrogen), a cost analysis proved
that cryomilling is an economically feasible
processing approach for the commercial fabrication of nanostructured materials.70 One of the
most interesting aspects of this approach is that
cryogenic temperatures make brittle the material
to be milled and this implies that the specific
energy required for milling is reduced. Additionally, cryogenic milling prevents the materials
from thermal damage, hinders the occurrence of
undesirable chemical reactions between phases70
and reduces particles aggregation.16 Obviously,
other particular reasons can suggest the use of
cryogenic milling. For example, Feng et al.71
found that cryogenic milling of griseofulvin
mainly implies drug crystallinity reduction due
to the increase of crystal defects, rather than the
formation of amorphous drug. This, in turn,
implies having a defective crystal whose bulk
properties differs from the amorphous form
(significant decrease of melting enthalpy as a
function of milling time but absence of glass
transition temperature). Jayasankar et al.,72
studying the reaction of co-crystal formation
during cogrinding, needed cryogenic conditions
to avoid that the reaction proceeded through the
melt phase forming in normal cogrinding.
In principle, many of the mills presented in
previous sections can be used for cryomilling
although some of them better adapt to cryogenic
conditions. Among them attritors, ball mills and
Pin mills can be remembered. In particular, a
widely used71,73,74 mill is SPEX model 6750. It
3974
consists of stainless steel vessel immersed in

liquid nitrogen, within which a stainless steel rod
was vibrated by means of a magnetic coil.
GRINDING MODELLING
Once practical evidences underline the effectiveness and the reliability of the mechanochemical
activation process, from both a theoretical and
industrial point of view, the necessity of process
optimisation arises. Obviously, this target can be
achieved by a classical trial and error procedure or
by the adoption of proper mathematical models
able to yield a mathematical metaphor (representation) of the entire process.25 While the first
approach is more convenient in the case of
particular mill types and considering small
variations of the operating conditions, the second
is to be preferred for the attainment of general
principles working for a wide range of operating
conditions and different mills. Regardless the
strategy adopted, however, the main question is
always the same: how do fixed operating conditions reflect on ground material properties? Or,
conversely: which are the operating conditions
leading to fixed ground material properties? In the
light of the trial and error approach, a valuable
help in answering to these specular questions is
given by the application of artificial neural
network (ANN). ANN is a theoreticalmathematical tool mimicking the learning processes of the
human brain. Indeed, it is constituted by elaboration units (ANN neurons or nodes) that are each
other interconnected.75 This means that, as real
neurons do, the elaboration unit receives information from and sends information to the other
elaboration units. On the basis of the interneurons connections, ANN assumes different
architectures.76 Feed forward ANN can be used
to predict output values after a proper training,
called learning, is performed. Briefly, ANN is
presented many input/output sets represented by
operating conditions (e.g. mill revolution speed,
charge and milling time) and the corresponding
ground material properties (e.g. mean particle
size and amorphous fraction). Accordingly, ANN
learns the relation between input and output
data. When new operating conditions are presented, ANN should be able to yield a reasonable
evaluation of output data (mean particle size and
amorphous fraction, in this example). Obviously,
the successful use of ANN strongly relies on the
quality and reliability of the learning step. One of
the most important advantages of ANN consists in

establishing a relation between input and output
data without the necessity of knowing the exact
mechanisms of the process allowing input data
transformation into output data. On the contrary,
the second approach (mathematical modelling) is
exactly aimed to the identification of the mechanisms leading from input to output data. Thus, in
order to answer to the two cardinal questions
above presented, we need to model mill dynamics
(enabling the evaluation of collisions frequency
and energy), how collision energy is transferred to
the ground material and, finally, how the energy
received modifies ground material properties in
terms of, for example mean particle size, amorphous fraction and so on.
Mill Dynamics
Tka c ova 7 shows a very interesting approach for
the estimation of the energy transferred to mill
charge working, in principle, for any kind of mill.
Starting point is the recognition that mechanical
energy is not continuously supplied to the charge
but it is discontinuously administered through
periodic, impulsive, events (grinding media
collisions). If t is the grinding duration, t1 is the
impulse duration and T is the impulse period, the
intrinsic grinding time tint is given by tint (t1/
T)t. Defining m and v as single grinding medium
mass and velocity, respectively, the associated
kinetics energy E is given by E (1/2)mv2.
Accordingly, the intrinsic mill power, Pint, can
be calculated as follows:
Pint
J
X
dE
j1
dt
J
X
mv
j1
J
dv X
Fv
dt
j1
(12)
where J, the number of impulses per unit time,

depends on mill operating conditions. Accordingly, mechanical energy W supplied to mill
charge is given by:
Wint Pint tint
J
X
j1
Fvt1
t FNt1 v
T
T
(13)
where N Jt is the number of all impulses. The

specific energy of grinding Wseg is given by Wint/M,
where M is mill charge mass (mass of the material
to be ground). On the basis of this analysis, Table 1
reports the expressions of N, v and Wseg for
different mills. According to this approach, balls
mills ranking according to Wseg sees planetary
DOI 10.1002/jps
3975
Table 1. Evaluation of Impulses Number N, Grinding Media Colliding Velocity v

and Specific Grinding Work Wseg for Different Mills
Mill
Planetary
Vibrational
Attritor
Jet
vt
vt
vt
N
v
p
2AD
4pva
pvD
vgas
Pheripheral speed
Tumbling
N
vt
pvD
p
2gD
Wseg
Wseg (Mgm/M)AvtD
Wseg (Mgm/2M)vt(4pva)2
Wseg (Mgm/M)vtgpDm(v)
Wseg N=2v2gas
Wseg (N/2)(pvD)2
Wseg (Mseg/2M)(pvtgpD)2
v is the rotational speed, t is the grinding time, g is the gravitational acceleration, A is the
acceleration, D is drum diameter, a is the amplitude, m(v) is the friction coefficient, Mgm is the
grinding media mass and M is the charge mass (adapted from Ref. 7).
mills in pole position, followed by vibrational,

attritors and tumbling mills.
Although Eqs. (12) and (13) represent very nice
and simple tools for a rough evaluation of the
energy transfer, a more precise evaluation of Wseg
requires a detailed modelling approach to mill
dynamics.77,78 For example some authors7981
apply the discrete element method (DEM) to
describe balls dynamic in tumbling mill. Interestingly, they find good agreement between
predictions and experimental observations for
what concerns balls movement and energy transfer. The analysis of planetary mill dynamics has
been undertaken by many researchers for its
large use in common practice. Particle element
method,82 analytical/numerical models83 and
DEM84 have been used to simulate balls dynamics
in planetary mills. Mio et al.84 found that specific
impact energy of balls increases with increasing
the ratio between vials rotation speed and circular
basement rotation speed, but it falls about the
critical speed ratio due to rolling motion. Among
the various attempts to model planetary mill
dynamics, it is worth mentioning that by Magini
and coworkers85 as it yields an analytical solution
for the power P
transferred to the unit mass of
charge:
P
b Nb mb tvp vn
3
vv Rv db =2
vp vv Rp

vp

Rv db =2

2pp
(14)
where p is the charge mass, t is the time, Rp is the

distance between the circular basement axis and
vial centre, Rv is the vial radius, vp is the circular
basement angular velocity, vv is the vial angular
DOI 10.1002/jps
velocity, mb is the grinding medium (ball) mass, db

is the grinding medium (ball) diameter, Nb is the
number of grinding media in each vial and wb is
the correcting factor depending on Nb. This model
allows discovering that energy transfer to mill
charge takes place both in the friction and in the
impact phase. Despite its simplicity, Eq. (14)
proved to be reliable.86 Castillo et al.69 successfully compared the calculated (Visual Nastran
software) dynamics of a vibrational mill with
the experimental detected one. In addition, he
evaluated grinding media collision frequency (N).
Wang,78 in his PhD thesis, built up a mathematical model simulating the 3D milling dynamics of
SPEX-8000 mills. Accordingly, he could evaluate
ball positions, velocities and impacts frequency. In
particular, he found that for different number of
balls, the frequency of impacts between balls and
the vial wall is proportional to the number of balls
and the number of impacts between balls is nearly
proportional to the square of the number of balls
and the square of radius of balls. Model simulations show good agreement with the results
coming from a probability analysis, indicating
that the model is reasonable. Always applying
DEM approach, Wang78 also considered attritors
modelling getting interesting preliminary results.
It is important to underline that the main problem
arising in ball mills modelling is the huge
computational duty due to the high number of
balls involved. We would like to finally mention an
interesting mathematical model devoted to
describe grinding in jet mills. Briefly, the authors
subdivided mill chamber in two zones: (1) grinding
and (2) central zone.87 An additional third zone is
represented by the external classificatory. Particles size reduction in zone 1 (grinding) is modelled
by considering the selectivity and the breakage
functions. The first one accounts for the particle
3976
fraction not undergoing diameter reduction while

the second one describes how particle size
modification occurs. The stochastic particles flow
from zone 1 to zone 2 (and vice versa) and from
zone 2 to zone 3 is described by three similar
probability functions. The model is completed by
two mass balances over zone 1 and 2 and referring
to each particles class. Model simulations led to
realistic results.
Energy Transfer to Mill Charge

Once mill dynamics has been properly accounted
for, the second important step to consider is how
mechanical energy is transferred to mill charge.
As, obviously, energy transfer strictly depends on
mill type, for their wide use, we can focus the
attention on ball mills. Among many other
valuable approaches, we believe that the one
proposed by Delogu and Cocco88 is very interesting. These authors, assuming that the mill charge
is kept homogeneous during grinding, affirmed
that energy transfer to mill charge is gradual and
progressive. Indeed, after the first collision, only a
small fraction k (typically of the order of 105
106) of the whole mill charge will be modified due
the collision energy. Thus, mill charge can be
subdivided into two classes, x0(1) and x1(1),
representing, respectively, the mill charge fraction that has never been impacted and the mill
charge fraction that underwent one impact after
n 1 impact. Obviously, x0 1 k and x1 k.
After the second impact, three mill charge classes
will exist: x0(2), x1(2) and x2(3). They represent,
respectively, the mill charge fraction that has
never been impacted and the fractions that
underwent, respectively, one and two collisions
after n 2 impacts. Obviously, charge classes
increases with n and this process continues
indefinitely until grinding stop. The authors
suggested the following relation for the evaluation
of each class variation (Dxi(n)) due to increasing
collisions number:
Dx0 n kx0 n;
Dxi n kxi n kxi1 n
(15)
8i > 0
Since the number of impacts occurring during a

milling process, n, is very large and the mass
fraction of powder processed in each collision, k, is
reasonably quite small, all the previous discrete
equations can be safely written in the following
continuous form:
dx0 n kx0 n dn;
(16)
dxi n kxi n dn kxi1 n dn

Their solutions read:
x0 n ekn ;
xi n
kni kn
e
i!
(17)
This means that x0 decreases exponentially with

n, while all the other classes xi increase, reach a
maximum (whose occurrence increases with i) and
then exponentially decrease (see Fig. 7). Once mill
dynamics is known, grinding media impact
frequency can be evaluated so that it is possible
to convert n into time. Accordingly, this approach
represents the connection between mill operating
condition, determining mill dynamics, and grinding time.
Energy Effect and Activation Yield

The final step regards the theoretical evaluation
of mechanical energy effect on mill charge.
Despite this is a key point of the entire grinding
process, the intrinsic difficulty of this topic put its
discussion out of the aim of this review.1,15,89 In
addition, in the light of the energetic cogrinding
Figure 7. According to the model proposed by Delogu

and Cocco,88 mill charge does not homogeneously
receive mechanical energy during milling. Accordingly,
he supposes that after n collisions, mill charge can be
subdivided into n classes. Class zero (x0), represents
never impacted mill charge fraction, class 1 (x1) represents mill charge fraction involved only in one collision,
class 2 (x2) represents mill charge fraction involved in
two collisions and so on up to class n (xn). According to
this theory, x0 decreases exponentially, while all other
classes show a maximum.
DOI 10.1002/jps
process optimisation, this knowledge is not

strictly required. Indeed, the energy balance
made up around the mill reads:
hE ET Ek Ea
(19)
where Ag and A represent, respectively, powder

surface area after and before grinding, while g sv
g
and gsv represent, respectively, powder surface
energy after and before grinding. Eab, instead, is
connected with bulk properties modifications
(namely, introduction of defects in the crystalline
network as discussed in Mechanical Activation
Section). Obviously, similar considerations can be
done for the cogrinding case, but the expression of
Eas is not equally straightforward as both the drug
and stabilising agent surface properties need to be
considered. Anyway, regardless of the grinding or
cogrinding situation, Eqs. (18) and (19) give the
opportunity of defining a global activation yield,
hga, as follows:
hga
Ea
Ea
h
E
ET Ek Ea
hE ET Ek
E
(21)
If hE can be easily evaluated on the basis of mill

engine nominal power, Ek estimation descends
from the knowledge of mill dynamics while ET
implies a mill thermal analysis. Although ET can
be theoretically estimated, it is usually more
convenient to build up a mill thermal model whose
DOI 10.1002/jps
APPLICATIONS
After the discussion about mechanical-activation
process and the tools needed for its realisation
(mills), it is interesting to present some results
descending from this approach. As drug activation
can be determined either directly on the coground
system by means of different techniques (see
Experimental Verification of Activation Section)
or it can be evaluated on the basis of its in vivo
effect (bioavailability enhancement), this section
is divided into two parts: in vitro and in vivo.
While the first part is essentially devoted to show
and discuss in vitro evidences of activation, the
second one deals with the in vivo evidences of
mechanical activation. In the impossibility of
giving an exhaustive presentation of all the many
examples regarding drug mechanochemical activation, but aimed to provide a rational presentation of them, in vitro applications are presented
according to the carrier used, while in vivo
applications are shown according to the pharmacological class of the drug (obviously, many other
possible criteria, such as milling type, energy
involved in milling and so on, could have been
adopted). Accordingly, in vitro section comprehends inorganic, polymeric and cyclodextrins
carriers. In vivo section considers anti-inflammatory, anti-tumoural, antihypertensive, antispasmodic and antifungal drugs.
In Vitro
(20)
As, in general, Ea is the most difficult term to be

estimated, it is convenient rewriting Eq. (20) in
the light of Eq. (18) which allows Ea expression as
a function of more convenient quantities (hE, ET
and Ek):
hga
fitting on mill temperature increase allows ET

determination.90 As in Eq. (18) E and ET are
comparable, we should expect low values for hga.
(18)
where h is the mill engine yield ( 0.97 for

electrical engine), hE is the energy effectively
supplied by the mill engine to the mill, ET, Ek and
Ea are, respectively, the thermally dissipated
energy (due to the attrition between mill moving
parts and colliding grinding media), the kinetic
energy owned by all the mill moving parts (in this
term it is comprised the potential energy associable with possible mill springs) and the energy
causing mill charge activation. In the case of one
component grinding, Ea can be subdivided into
two terms: Eas and Eab. Eas is connected with the
increase of surface area (comminution) and surface properties modifications. Accordingly, we
have:
sv
Ea Eas Eab Ag g sv
g Ag Eab
3977
Inorganic Carriers
Examples of inorganic carriers are calcium
silicate and silicon dioxide.91 For example, Bahl
and Bogner91 studied the indomethacin (g-polymorph) (IM) activation process recurring to its
(low energy) cogrinding with Neusilin US2
(amorphous magnesium aluminometasilicate,
specific surface are 300 m2/g) in a rolling jar
mill consisting of a cylindrical porcelain jar
(internal volume 1000 mL) hosting zirconia balls.
Different IM-Neusilin US2 weight ratios were
considered (1:5, 1:4, 1.1, 1:0.5). Interestingly,
these authors found that, whatever the drug
carrier ratio considered, the relative humidity
(RH) of the cogrinding environment highly
3978
influences amorphisation kinetics. For example

when RH 75% and IM:Neusilin US2 1:5, 60%
of the whole drug was converted into amorphous
phase after 0.5 days while similar conversions for
RH 0% and the same weight ratio occurred after
2 days. Watanabe et al.92 used a vibrational mill
made up by a cylindrical zirconia bowl (100 mL)
hosting 74 zirconia balls (diameter 10 mm) to
get IM (g-polymorph) amorphisation. Four different carriers were used: talc (3MgO4SiO2H2O),
SiO2, Mg(OH)2 and a 0.42:0.58 (w/w) Mg(OH)2:SiO2 mixture (PMS). The drug:carrier weight
ratio was constantly equal to 1:1. The neatly
higher energetic process used by Watanabe in
comparison to that of Bahl and Bogner91 made
possible an almost complete amorphisation in
10 min in the case of IMPMS system. IMSiO2
and IMtalc systems yielded to a complete
amorphisation after 60 and 30 min, respectively.
On the contrary, IMMg(OH)2 system did not
yield a complete amorphisation in the maximum
experimental time range considered (60 min).
According to the authors, this behaviour was due
to optimal interaction of IM with talc or SiO2.
Shakhtshneider et al.93 employed a planetary mill
(vial volume 40 mL; ball diameter 6 mm;
grinding media:mill charge 20:1 (w/w)) to study
the activation process of Ibuprofen (IB) in
presence of talc (1:10, w/w). XRPD and DSC
analysis demonstrated that the authors could
obtain complete drug amorphisation. Interestingly, dissolution studies (a known amount of
coground material was put in 100 mL water at
378C under mixing) revealed that the coground
system yielded, in about 30 min, to a IB
concentration of 500 mg/mL which is approxi-
mately 12.5 times native IB solubility in water

(see Tab. 2).
Polymeric Carriers
Although different polymeric carriers can be used
(dextrans, chitin, chitosan, gelatin, polyethylene
glycol, methyl cellulose, hydroxypropyl cellulose91), polyvinylpyrrolidone (PVP) is one of the
most used. For example, Shakhtshneider et al.93
employed this carrier to activate sulfathiazole and
Piroxicam adopting the same milling conditions
above reported for Ibuprofen activation with talc.
In the case of sulfathiazole, they found that part of
the drug was still crystalline after cogrinding
when the drug:PVP weight ratio is 3:1 or 1:1
(regardless milling time; 4, 6, 8 or 12 min), while
the drug was completely amorphous when this
ratio was 1:3 (12 min cogrinding). Dissolution
studies performed on the 1:3 system (12 min
cogrinding) yielded to a drug concentration of
7000 mg/mL after 1 h. This concentration is
approximately 10 times the water solubility of
native sulfathiazole (see Tab. 2). In the case of
Piroxicam, the increase of apparent solubility was
approximately 2.5 and 4 times that of the
native drug when the drug:PVP weight ratio
was equal to 1:1 and 1:10, respectively. Obviously,
dissolution studies performed on simple physical
drugPVP mixtures (these systems did not
undergo cogrinding), did not imply any significant
improvement of drug solubility. Watanabe et al.92
used PVP to activate Indometacin (IM; g-polymorph) (IM:PVP 1:1, molar ratio) in a vibrational mill (cogrinding times: 30, 60, 120 and
180 min). XRPD revealed that IM diffraction
Table 2. Drugs Solubility in Aqueous Environment

Drug
Class
Glibenclamide
Gliquidone
Glisentide
Anti-diabetes
Hypoglycaemic agent
Anti-diabetes
Griseofulvin
Ibuprofen
Indomethacin
Methylhydroxyprogesterone
Nifedipine
Nimesulide
Sulfathiazole
Antifungal
Nonsteroidal anti-inflammatory
Low dose: progestinic activity
High doses: anticancer
Calcium-channel blocker
Antimicrobial
Solubility
References
0.3 mg/cm (258C, water)

0.14 mg/cm3 (258C, water)
1.5 mg/cm3 (378C, artificial gastric
medium without enzymes)
11.9 mg/cm3 (378C, water)
40 mg/cm3 (378C, water)
1.2 mg/cm3 (378C, water pH 5.5)
98
97
94
25
93
105
31

12 mg/cm3 (378C, water, pH 5.5)
103
100
93
DOI 10.1002/jps
peaks decreased with increasing cogrinding time

to totally disappear after 180 min (totally amorphous IM). For this last system, apparent IM
solubility in water (378C) was raised up to 77 mg/
mL. Mura et al.94 studied glisentide (antidiabetic)
activation with PVP in a high energy vibrational
mill for different milling times (15180 min) and
drug:carrier weight ratios (1:3, 1:1 and 3:1). DSC
and XRPD confirmed that glisentide complete
amorphisation occurred after only 30 min in 1:3
systems. Interestingly, dissolution tests, performed at 378C in 1000 mL of artificial gastric
medium without enzymes, indicated that after
60 min glisentide concentration in the release
environment was around 5 mg/mL while, after the
same time, the 1:3 physical mixture yielded to
drug concentration of about 1.5 mg/mL. Shakhtshneider et al.95,96 studied indomethacin (IM) and
piroxicam activation by cogrinding with PVP in a
cryogenic mill (6750 Freezer/Mill, Inc., Metuchen,
New Jersey). 1 g sample (PVP/drug mass/mass
ratio ranging from 0.1 to 0.8) was milled at an
impact frequency of 10 Hz alternating milling
periods of 2 min with 1 min cool-down period
(overall milling time spanned between 60 and
78 min). The authors were interested in evaluating the stabilisation action of PVP by comparing
re-crystallisation kinetics relative to pure ground
drugs and to coground drugs. The necessity of
recurring to cryo-milling was dictated by the
impossibility of getting a complete amorphous by
simple drugs milling at room temperature in the
absence of the stabilising carrier. They found that
PVP exerts a good stabilising action on both drugs
even if better results were evidenced for IM.
Obviously, for both drugs, PVP stabilising action
increased in reason of its content in the coground
mixture. Finally, they found that coground cryomilled IM dissolution was neatly improved with
respect that of the same mixture that did not
undergo cryo-milling process.
Cyclodextrins
For their chemical and physical peculiarities,
cyclodextrins are widely used as stabilising agents
(carrier) in drug mechanochemical activation.
Indeed, a, b, g and substituted cyclodextrins
can be used. For example, Miro et al.97 coground
gliquidone (hypoglycaemic agent) with hydroxypropyl-b-cyclodextrin (HPbCD) in 1:2 molar ratio.
DSC and XRPD analyses revealed that the
authors obtained a system showing a very small
content of original crystalline drug, being the drug
DOI 10.1002/jps
3979
majority in the amorphous state. Dissolution

studies, performed at 378C in 1000 mL of 0.1 M
phosphate buffer (pH 7.4), revealed that, after
60 min, drug concentration in the release environment was around 9 mg/mL, that is approximately, 64 times drug solubility in the same
environment and conditions (see Tab. 2). Interestingly, oral administration of this coground
system (rats, drug dose 300 mg/kg) caused a
reduction of plasma glucose concentration that
was approximately 1.52 times that of native
gliquidone in the first 15 h following administration. Fukami et al.98 focused the attention
on glibenclamide (anti-diabetes) bioavailability
enhancement recurring to cogrinding with highly
branched cyclodextrins (HBCD). HBCD is a cyclic
glucan produced from waxy corn starch by the
cyclisation reaction of branching enzyme. Ball
mill with drug:HBCD 1:5 weight ratio was
considered. Grinding time was fixed in 2 h and
150 rpm. Interestingly, DSC analysis revealed
that glibenclamide melting point in the coground
system was reduced of 5.48C respect to the native
drug that melts at 170.18C. This was the proof that
the original crystalline drug was completely
disappeared in favour of nano-crystalline one.
The authors verified that cogrinding increased
drug apparent solubility up to 12.4 mg/mL, being
native drug solubility equal to 0.3 mg/mL.
These examples clearly show that, apart from
the specific energy supplied by the mill, key
factors for the attainment of drug activation are
drug:carrier ratios >1:2 and good chemico-physical interactions between drug and carrier.
In Vivo
While previous section was focussed on the in vitro
evaluation of coground systems activation, this
section shows the in vivo evidences of activation
and their relations with in vitro release tests.
Indeed, what is often desired is to get information
about in vivo performance resorting to in vitro
behaviour. In order to rationalize the presentation, examples are subdivided according to drug
pharmacological class.
Anti-Inflammatory Drug
Perret and Venkatesh55 showed data referring to
an activated system composed by drug X (antiinflammatory, poorly water-soluble crystalline
drug) and crosslinked polyvinylpyrrolidone
3980
(PVPclm) in a weightweight ratio equal to 1:2.

The system, once coground in a vibrational mill,
was characterised by 40% amorphous drug and
60% nanocrystalline drug (original crystalline
drug absent). Figure 8A clearly shows the better
release performance of the activated system (filled
circles) in comparison to that of an identical w/w
ratio drug-polymer physical mixture (open circles). While after about 10 min physical mixture
yields to a release environment concentration just
below 2 mg/mL, the activated system concentration is up to 8.5 mg/mL after about 50 s and then,
due to amorphous drug re-crystallisation, concentration decreases to 7.9 mg/mL. Interestingly,
the evident in vitro superiority of the activated
Figure 8. (A) In vitro test referring to drug X (poor

water soluble drug) release from activated drug X crosslinked polyvinylpyrrolidone (1:2 (w/w) ratio) system
(filled circles). Forty percent of the original drug is
amorphous while the remaining 60% is nanocrystalline.
Open circles indicate the release behaviour of an equal
w/w ratio physical mixture that did not undergo
cogrinding (activation). C is release environment drug
concentration, while t is time. (B) Comparison between
in vivo behaviour referring to the activated system
(filled circles) and a commercial reference (open circles).
C is blood concentration while t is time.
system reflects into a similar in vivo superiority

as shown in Figure 8B. This figure reports drug X
plasma concentration (humans) following oral
administration of a 200 mg dose commercial
reference (open circles) and activated (filled
circles) tablet made by the same activated system
tested in vitro (Fig. 8A). Not only the maximum
blood concentration (Cmax) relative to the activated system is approximately two times that of
the reference, but the area under the curve (AUC)
relative to the activated system is approximately
1.3 times that of the reference. It is, thus, evident
the consistent increase of drug X bioavailability.
Magarotto and coworkers99 presented data
regarding the activation of nimesulide, a low
water soluble ( 100 mg/mL, pH 7.5, 378C,100
melting temperature 148.78C) nonsteroidal anti
inflammatory drug. Cogrinding, performed in a
vibrational mill using b-cyclodextrin (bCD) as
carrier (nimesulide:bCD w/w ratio is 1:3), led to
the complete transformation of the original drug
crystals into nano-crystals melting at 145.78C.
This means that nano-crystals were characterised
by a radius of approximately 23 nm (see Fig. 4).
The in vivo behaviour (see Fig. 9) of the activated
system (filled circles), showed a little improvement with respect to a commercial reference
(open circles). Indeed, Cmax was increased from
3.3 mg/mL (commercial reference) to 3.6 mg/mL
(coground system) and AUC passed from 20.4 mg
h/mL (commercial reference) to 22.4 mg h/mL
(coground system). The little improvement of the
in vivo performance of the coground system was
coherent with its low degree of activation as no
Figure 9. Comparison between in vivo behaviour

referring to activated nimesulideb-cyclodextrin (1:3
(w/w) ratio) system (filled circles) and a commercial
reference (open circles). C is blood concentration while
t is time.
DOI 10.1002/jps
3981
amorphous drug was present and nano-crystals

were characterised by a big radius. Nevertheless,
this low activation determined an improvement
for what concerns the onset of action.
Anti-Tumoural Drugs
Carli et al.101 focussed the attention on the
activation of methylhydroxyprogesterone acetate
(MPA), a crystalline drug showing low water
solubility (1.2 mg/mL at pH 5.5 and 378C), scarce
water wettability (water contact angle 838180 )
and high melting temperature (2068C) and
enthalpy (88 J/g). While at low doses MPA exerts
a progestinic activity, at high doses presents
anticancer properties. These authors chose bcyclodextrin (bCD) as carrier and made use of a
planetary mill to cogrind MPA and bCD in a molar
ratio equal to 1:2. As neither original crystalline
drug nor drug nano-crystals could be found in the
in the coground system, only amorphous MPA was
present in the activated system. Figure 10A shows
the comparison between the release kinetics
(dispersed amount method: excess amount of
activated system was put in 50 mL buffer solution,
pH 5.5, 378C) pertaining to the activated system
(filled circles) and the physical mixture characterised by the same MPA:bCD molar ratio (open
circles). The advantage gained by using the
activated system is evident. It is worth mentioning that the increase in MPA release kinetics from
the activated system was due to both the
amorphous MPA condition and to the increased
coground system wettability. Indeed, the water
contact angle of the coground system was lowered
to 308580 , approximately one-third of the pure
MPA one. Also in this case, the promising
behaviour shown in vitro reflects into positive
evidences in vivo. Figure 10B shows MPA plasma
concentration (beagle dogs) following oral administration of two 50 mg dose capsules containing
the activated system (filled circles) and a 250 mg
dose of a standard commercial tablet (open circles)
containing original crystalline MPA. The superiority of the activated system performance relies
in the higher Cmax (40 ng/mL, commercial tablet;
337 ng/mL, activated system) and AUC (485 ng h/
mL, commercial tablet; 1526 ng h/mL, activated
system).
Antihypertensive Drugs
Sugimoto et al.102 led an extensive study on
bioavailability enhancement of Nifedipine (low
water soluble ( 5 mg/mL, pH 5.5, 308C103)
DOI 10.1002/jps
Figure 10. (A) In vitro test referring to methylhydroxyprogesterone (MPA) release from activated
MPAb-cyclodextrin (1:2 molar ratio) system (filled
circles). All the original drug is amorphous. Open circles
indicate the release behaviour of an equal molar ratio
physical mixture that did not undergo cogrinding (activation). C is release environment drug concentration,
while t is time. (B) Comparison between in vivo behaviour referring to the activated system (filled circles)
and the same molar ratio physical mixture (open circles). C is blood concentration while t is time.
antihypertensive drug) by means of cogrinding

with polyethylene glycol 6000 (PEG 6000) and
hydroxypropylmethyl cellulose (HPMC). Five
hundred milligrams of nifedipine and 2.5 g of
HPMC were added to 500 mg of molten PEG 6000
(708C). This mixture was let to solidify at room
temperature and then coground in a spex mill for
10 min. The coground system was added with
3 mL water and then coground again for further
10 min. After cogrinding, water was removed by
drying at 708C for 6 h. XRPD analysis revealed
that a small percentage of original crystalline
drug still existed in the coground system.
Figure 11A shows the in vitro nifedipine release
3982
huge increase of Cmax (89 ng/mL, coground

system; 9.1 ng/mL, physical mixture) and AUC
(122 ng h/mL, coground system; 47 ng h/mL,
physical mixture) proved the neat bioavailability
improvement due to activation. Interestingly, the
authors noted that if water was not added to the
ternary system, drug activation was not equally
pronounced.
Antispasmodic Drugs
Sawayanagi et al.104 studied the mechanochemical activation of phenytoin by cogrinding it
with chitin or chitosan. They found that a 1:2
Phenytoin: chitosan coground mixture orally
administered to beagle dogs comported a substantial increase in phenytoin blood concentration. In particular, the AUC corresponding to
activated drug was approximately three times
that of native phenytoin and that maximum drug
concentration was roughly 3.2 times that competing to the native drug. Similar encouraging
results were also found for mechanochemical
activation of phenytoin in presence of cellulose
(1:9 drug:polymer ratio). Interestingly, in vitro
dissolution tests indicated a neatly higher dissolution rate in comparison with that of the native
drug.16
Figure 11. (A) In vitro test referring to nifedipine
release from activated nifedipinepolyethylene glycol
6000hydroxypropylmethyl cellulose (1:1:5 (w/w) ratio)
system (filled circles). Almost all the original drug is
amorphous. Open circles indicate the release behaviour
of an equal w/w ratio physical mixture that did not
undergo cogrinding (activation). t is time. (B) Comparison between in vivo behaviour referring to the activated system (filled circles) and the same w/w ratio
physical mixture (open circles). C is blood concentration
while t is time.
(JP paddle method, 900 mL water, 378C, stirring

100 rpm) referring to the activated system (filled
circles) and the identical physical mixture (open
circles). Also in this case, the better performance
of the activated system results evident. This
interesting in vitro behaviour reflected into an
analogously interesting in vivo behaviour as
witnessed by Figure 11B. This figure shows
nifedipine plasma concentration (beagle dogs)
following oral administration (by means of a
sonde) of 50 mL water suspension containing an
amount of activated system (filled circles) or the
same amount of physical mixture (open circles)
corresponding to a dose of 10 mg nifedipine. The
Antifungal Drugs
The mechanochemical activation of griseofulvin in
mixture with different polymeric carriers (potato
starch, polyethylene glycol, chitosan) led to
improved in vitro dissolution characteristic with
respect to that of the native drug.16 This promising in vitro behaviour reflected in good in vivo
results. Indeed, experiments conducted on rats
and rabbits, demonstrated an increase of bioavailability equal to 40% and 50%, respectively
(vibrational mill, drug/polymer ratio ranging
from 1:2 to 2.1). Interestingly, urinary excretion
tests allowed verifying that the increased
griseofulvin concentration was due to an increase
of the (pharmacokinetic) absorption constant as
the (pharmacokinetic) elimination constant was
substantially unaffected by mechanochemical
activation.
CONCLUSIONS
Mechanochemical activation represents a
versatile and valuable tool to improve drug
bioavailability. The versatile aspect relies in
DOI 10.1002/jps
the possibility of properly tuning the activation

conditions such as carrier type, drug:carrier ratio,
the possibility of contemporarily using more than
one carrier, milling time and milling type
(determining the amount and rate of energy
transfer to the coground system) and the possibility of using water to favour the activation
process. The valuable aspect consists in the huge
drug bioavailability improvement that this
approach can ensure without recurring to the
use of solvents (other than water) whose elimination from the final form can be very problematic in
relation, for example, to admissible residue. Other
advantages of this approach are that it does not
lead to new chemical entities whose presence
could be cause of regulatory problems and that it
does not require expensive instrumentation. In
addition, the development of theoretical studies
about mechanochemical activation ensures a
global understanding of the whole process, this
being of paramount importance in the designing
and optimisation of the entire process. Obviously,
mechanochemical activation does not represent
an approach good for all seasons as it works only
for class II drugs36, that is drugs showing
low water solubility and good gastro-intestinal
permeability.
ACKNOWLEDGMENTS
This work was supported by the Fondazione
Cassa di Risparmio of Trieste, by Fondazione
per il Sostegno delle Strutture Cardiovascolari,
Mirano, VE, by the Fondazione Casali of Trieste
and Fondo Trieste 2006.
REFERENCES
1. Tka c ova K. 1993. First international conference on
mechanochemistry: An introduction. Proc. First
Intl. Conf. on Mechanochemistry Kos ice, Slovak
Republic, Vol. 1: pp. 917.
2. Boldyrev VV, Tka c ova K. 2000. Mechanochemistry
of solids: Past, present and prospects. J Mat Synth
Processing 8:121132.
3. Ostwald W. 1919. Hanbuch der allgemeinnen chemie. Leipzig: Akad Verlagsantalt.
4. Suryanarayana C, editor. 1999. Non-equilibrium
processing of materials. Oxford: Pergamon.
5. Heinicke G. 1984. Tribochemistry. Berlin: Akademie-Verlag.
DOI 10.1002/jps
3983
6. Butyagin PY. 1989. Active states in mechanochemical reactions. Glasgow: Hardwood Acad. Publ.
7. Tka c ova K. 1989. Mechanical activation of
minerals. Amsterdam: Elsevier.
8. Juhasz AZ, Opoczky L. 1990. Mechanical activation of minerals by grinding. Budapest: Akademiai
Kiado.
9. Shingo PH, editor. 1992. Mechanical alloying. Zu rich: Trans. Tech. Publications.
10. Gutman EM. 1994. Mechanochemistry of solid
surfaces. London: World Scientific.
11. Yalkowsky SH. 1981. Techniques of solubilisation
of drugs. New York: Marcel Dekker.
12. Boldyrev VV. 2004. Mechanochemical modification and synthesis of drugs. J Mater Sci 39:
51175120.
13. Lovrecich ML. 1995. Supported drugs with
increased dissolution rate and process for their
preparation. US Patent 5449521.
14. Dobetti L, Bresciani M. 2003. Process for activation of drugs in a vibrational mill. European patent
1406728.
15. Boldyrev VV. 1993. Mechanical activation of
solids. Proc. First Intl. Conf. on Mechanochemistry
Kos ice, Slovak Republic, Vol. 1: pp. 1826.
16. Shakhtshneider TP, Boldyrev VV. 1999. Mechanochemical synthesis and mechanical activation
of drugs. In: Boldyreva E, Boldyrev VV, editors.
Reactivity of molecular solids. John Wiley & Sons
Ltd. Chichester, GB.
17. Tse JS. 1992. Mechanical instability in ice Ih. A
mechanism for pressure-indiced amorphization.
J Chem Phys 96:54825487.
18. Johnson WL, Li M, Krill CE. 1993. The crystal to
glass transformation in relation to melting. J NonCryst Solids 156158:481492.
19. Boldyrev VV. 1996. Mechanochemistry and
mechanical activation. Mater Sci Forum 225
227: 511520.
20. Hu ttenrauch R, Fricke S, Zielke P. 1985. Mechanical activation of pharmaceutical systems. Pharm
Res 2:302306.
21. Grant DJW, York GP. 1986. Entropy of processing:
A new quantity for comparing the solid state disorder of pharmaceutical materials. Int J Pharm
30:161180.
22. Amin AE, Ahlneck C, Alderborn G, Nystro m C.
1994. Increased metastable solubility of milled
griseofulvin, depending on the formation of a disordered surface structure. Int J Pharm 111:159
170.
23. Byrn SR, Pfeiffer RR, Stephenson G, Grant DJW,
Gleason WB. 1994. Solid-state pharmaceutical
chemistry. Chem Mater 6:11481158.
24. Tipper CF. 1962. The brittle fracture story. London: Cambridge University Press.
25. Grassi M, Grassi G, Lapasin R, Colombo I. 2007.
Understanding drug release and absorption
3984
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
mechanisms: A physical and mathematical

approach. Boca Raton: CRC Press.
Freundlich H. 1923. Colloid and capillary
chemistry. New York: E.P. Dutton and Company
Inc.
Adamson AW, Gast AP. 1997. Physical chemistry
of surfaces. New York: Wiley-Interscience Publications.
Kaya E, Hogg R, Kumar SR. 2002. Particle shape
modification in comminution. Kona 20:188195.
Buckton G, Beezer AE. 1992. The relationship
between particle size and solubility. Int J Pharm
82:R7R10.
Bikerman JJ. 1970. Physical surfaces. New York:
Academic Press.
Grassi M, Colombo I, Lapasin R. 2000. Drug
release from an ensemble of swellable crosslinked
polymer particles. J Control Rel 68:97113.
Grassi M, Coceani N, Magarotto L, Ceschia D.
2003. Effect of milling time on release kinetics
from co-ground drug polymer systems. Proc.
2003 AAPS Annual Meeting and Exposition, Salt
lake City, # M1201.
Madras G, McCoy BJ. 2003. Growth and ripening
kinetics of crystalline polymorphs. Cryst Growth
Des 3:981990.
Stephenson GA, Forbes RA, Reutzel-Edens SM.
2001. Characterisation of the solid state: Quantitative issues. Adv Drug Del Rev 48:6790.
Bugay DE. 2001. Characterisation of the solidstate: Spectroscopic techniques. Adv Drug Del
Rev 48:4365.
Yu L. 2001. Amorphous pharmaceutical solids:
Preparation, characterization and stabilization.
Adv Drug Del Rev 48:2742.
Zhang M, Efremov MY, Schiettekatte F, Olson EA,
Kwan AT, Lai SL, Wisleder T, Greene JE, Allen
LH. 2000. Size-dependent melting point depression of nanostructures: nanocalorimetric measurements. Phys Rev B 62:1054810557.
Huang WJ, Sun R, Tao J, Menard LD, Nuzzo RG,
Zuo JM. 2008. Coordination-dependent surface
atomic contraction in nanocrystals revealed by
coherent diffraction. Nat Mater 7:308313.
Lu K, Jin ZH. 2001. Melting and superheating of
low-dimensional materials. Curr Opin Solid Mater
Sci 5:3950.
Zhao M, Zhou XH, Jiang Q. 2001. Comparison of
different models for melting point change of metallic nanocrystals. J Mater Res 16:33043319.
Sanz N, Boudet A, Ibanez A. 2002. Melting behavior of organic nanocrystals grown in sol-gel
matrices. J Nanop Res 4:99113.
Jackson CL, McKenna GB. 1990. The melting
behavior of organic materials confined in porous
solids. J Chem Phys 93:90029018.
Brun M, Lallemand A, Quinson JF, Eyraud C.
1973. Changement detat liquide-solide dans les
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
milieux poreux. II Etude theorique de la solidification dun condensat capillaire. J Chim Phys 70:
979989.
Magomedov MN. 2004. Dependence of the surface
energy on the size and shape of a nanocrystal.
Phys Solid State 46:954968.
Liu F, Yang G, Wang H, Chen Z, Zhou Y. 2006.
Nano-scale grain growth kinetics. Thermochim
Acta 443:212216.
Skrdla PJ. 2007. Crystallization of glycine during
freezing of a 40/60 w/w sucrose/glycine excipient
system: An alternative to the JohnsonMehl
Avrami (JMA) equation for modeling dispersive
kinetics. J Pharm Sci 96:21072110.
Bhugra C, Shmeis R, Krill SL, Pikal MJ. 2006.
Predictions of onset of crystallization from experimental relaxation times. I. Correlation of molecular mobility from temperatures above the glass
transition to temperatures below the glass transition. Pharm Res 23:22772290.
Bergese P, Colombo I, Gervasoni D, Depero LE.
2003. Assessment of the X-ray diffraction-absorption method for quantitative analysis of largely
amorphous pharmaceutical composites. J Appl
Cryst 36:7479.
Proffen Th, Billinge SJL, Egami T, Louca D. 2003.
Structural analysis of complex materials using the
atomic pair distribution functionA practical
guide. Z Kristallogr 218:132143.
Lutterotti L, Ceccato R, Dal Maschio R, Pagani E.
1998. Quantitative analysis of silicate glass in
ceramic materials by the Rietveld method. Mater
Sci Forum 278281:8792.
Nogami H, Nagai T, Yotsuyanagi T. 1969. Dissolution phenomena of organic medicinals involving
simultaneous phase changes. Chem Pharm Bull
17:499509.
Pharmacos 4, Eudralex collection, medicinal products for human use: Guidelines, volume 3C, p.
234 (Internet site: http://pharmacos.eudra.org/F2/
eudralex/vol-3/home.htm).
Yu LX, Gatlin L, Amidon GL. 2000. Predicting oral
drug absorption. In: Amidon GL, Lee PI, Topp EM,
editors. Transport processes in pharmaceutical
systems. New York: Marcel Dekker. pp. 377409.
Amidon GL, Lennerna s H, Shah VP, Crison JR.
1995. A theoretical basis for a biopharmaceutic
drug classification: The correlation of in vitro drug
product dissolution and in vivo bioavailability.
Pharm Res 12:413420.
Perrett S, Venkatesh G. 2006. Enhancing the
bioavailability of insoluble drug compounds. Innov
Pharm Technol 19:8085.
Charman SA, Charman WN, Rogge MC, Wilson
TD, Dutko FJ, Pouton CW. 1992. Self-emulsifying
drug delivery systems: Formulation and biopharmaceutic evaluation of an investigational lipophilic compound. Pharm Res 9:8793.
DOI 10.1002/jps
57. Constantinides PP. 1995. Lipid microemulsions

for improving drug dissolution and oral absorption: Physical and biopharmaceutical aspects.
Pharm Res 12:15611572.
58. Gasco MR. 1997. Microemulsions in the pharmaceutical field: Perspectives and applications. In:
Solans C, Kunieda H, editors. Industrial applications of microemulsions. New York: Marcel Dekker. pp. 97122.
59. Pouton CW. Formulation of self-emulsifying drug
delivery systems. Adv Drug Del Rev 25:4758.
60. Kumar P, Mittal LK. 1999. Handbook of microemulsion science and technology. New York: Marcel Dekker. pp. 755771.
61. Ralza G. 2005. Modifiche delle proprieta` chimicofisiche di un principio attivo mediante miscele
binarie con ciclodestrine. Graduate thesis, Dept.
of Chem. Eng., Trieste University, Trieste, Italy.
62. Dobetti L, Cadelli G, Furlani D, Zotti M, Ceschia
D, Grassi M. 2001. Reliable experimental setup for
the drug release from a polymeric powder loaded
by means of co-grinding. Proc 28th Int Symp Controlled Release Bioact Mater, pp. 728729.
63. Roche USA, Invirase1 Sanquinavir Mesylate Capsules and Tablets Prescribing Information.
64. Van Mourik ID, Thomson M, Kelly DA. 1999.
Comparison of pharmacokinetics of Neoral1 and
Sandimmune1 in stable paediatric liver transplant recipients. Liver Transpl Surg 5:107
111.
65. Ueda Y, Shimojo F, Shimazaki Y, Kado K, Honbo
T. Solid Dispersion Composition of FR-900506
Substance. US Patent 4916138.
66. Gilis PMV, De Conde VFV, Vandecruys RPG.
Beads having a core coated with an antifungal
and a polymer. US Patent 5633015.
67. Tka ova K, Heegn H, Tevulova . N. 1993. Energy
transfer and conversion during comminution and
mechanical activation. Int J Miner Process 40:
1731.
68. Suryanarayana C. 2001. Mechanical alloying and
milling. Prog Mater Sci 46:1184.
69. Castillo J, Coceani N, Grassi M. 2004. Theoretical
and experimental investigation on the dynamics of
a vibrational mill. Convegno GRICU 2004, Nuove
Frontiere di Applicazione delle Metodologie dellIngegneria Chimica, Porto dIschia (Napoli). Vol.
I: pp. 165168.
70. Ye J, Schoenung JM. 2004. Technical cost modelling for the mechanical milling at cryogenic temperature (cryomilling). Adv Eng Mater 6:656
664.
71. Feng T, Pinal R, Carvajal MT. 2008. Process
induced disorder in crystalline materials: Differentiating defective crystals from the amorphous
form of griseofulvin. J Pharm Sci 8:32073221.
72. Jayasankar A, Somwangthanaroj A, Shao ZJ,
Rodrguez-Hornedo N. 2006. Cocrystal formation
DOI 10.1002/jps
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
3985
during cogrinding and storage is mediated by

amorphous phase. Pharm Res 23:23812392.
Crowley KJ, Zografi G. 2002. Cryogenic grinding of
indomethacin polymorphs and solvates: Assessment of amorphous phase formation and amorphous phase physical stability. J Pharm Sci 91:
492507.
Lolo M, Pedreira S, Va zquez BI, Franco CM,
Cepeda A, Fente CA. 2007. Cryogenic grinding
pre-treatment improves extraction efficiency of
fluoroquinolones for HPLC-MS/MS determination
in animal tissue. Anal Bioanal Chem 387:1933
1937.
Yamashita F, Takayama K. 2003. Artificial neural
network modeling for pharmaceutical research.
Adv Drug Del Rev 55:11171127.
Baughman DR, Liu YA. 1995. Neural networks in
bioprocessing and chemical engineering. San
Diego: Academic Pres.
Ajaol TT. 1999. The Development and Characterization of a Bali Mill for Mechanical Alloying.
Ph.D. thesis, Department of Materials and Metallurgkal Engineering, Queens University, Kingston, Ontario, Canada.
Wang W. 2000. Modelling and simulation of
dynamic process in high energy ball milling of
metal powders. Ph.D. thesis, Materials and Processing Engineering Dept., The University of
Waiakato.
Venugopal R, Rajamani RK. 2001. 3D simulation
of charge motion in tumbling mills by the discrete
element method. Powder Technol 115:157166.
Rajamani RK, Mishra BK, Venugopal R, Datta A.
2000. Discrete element analysis of tumbling. Powder Technol 109:105112.
Mishra BK. 2003. Review of computer simulation
of tumbling mills by the discrete element method
Part II-Practical applications. Int J Miner Process
71:95112.
Kano J, Saito F. 1998. Correlation of powder characteristics of talc during planetary ball milling
with the impact energy of the balls simulated by
the particle element method. Powder Technol 98:
166170.
Chattopadhyay PP, Manna I, Talapatra S, Pabi
SK. 2001. A mathematical analysis of milling
mechanics in a planetary ball mill. Mater Chem
Phys 68:8594.
Mio H, Kano J, Saito F, Kaneko K. 2002. Effects of
rotational direction and rotation-to-revolution
speed ratio in planetary ball milling. Mater Sci
Eng A332:7580.
Burgio N, Iasonna A, Magini M, Martelli S,
Padella F. 1991. Mechanical alloying of the FeZr system. Correlation between input energy and
end products. Il Nuovo Cimento 13D:459476.
Magini M, Colella C, Guo W, Martelli S, Padella F.
1994. Some hints about energy transfer in the
3986
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
mechanosynthesis of materials. Int J Mechanochem Mech Alloying 1:1425.

Gommeren HJC, Heitzmann DA, Kramer HJM,
Heiskanen K, Scarlett B. 1996. Dynamic modelling of a closed loop jet mill. Int J Miner Process
4445:497506.
Delogu F, Cocco G. 2000. Relating single-impact
events to macrokinetic features in mechanical
alloying processes. J Mater Synth Processing 8:
271277.
Butyagin PY. 1993. First international conference
on mechanochemistry: An introduction. Proc.
First Intl. Conf. on Mechanochemistry, Kos ice,
Slovak Republic, Vol. 1: pp. 2733.
Castillo JM, Lapasin R, Grassi M. 2003. Thermal
modelling of a vibrational mill. Ind Eng Chem Res
42:20152021.
Bahl D, Bogner RH. 2006. Amorphization of indomethacin by co-grinding with neusilin US2: Amorphization kinetics. Phys Stab Mech Pharm Res
23:23172325.
Watanabe T, Hasegawa S, Wakiyama N, Kusai A,
Senna M. 2003. Comparison between polyvinylpyrrolidone and silica nanoparticles as carriers for
indomethacin in a solid state dispersion. Int
J Pharm 250:283286.
Shakhtshneider TP, Vasilchenko MA, Politov AA,
Boldyrev VV. 1997. Mechanical preparation of
drug-carrier solid dispersions. J Therm Anal 48:
491501.
Mura P, Cirri M, Faucci MT, Gine`s-Dorado JM,
Bettinetti GP. 2002. Investigation of the effects of
grinding and co-grinding on physicochemical properties of glisentide. J Pharm Biomed Anal 30:227
237.
Shakhtshneider TP, Dane`de F, Capet F, Willart
JF, Descamps M, Myz SA, Boldyreva E, Boldyrev
VV. 2007. Grinding of drugs with pharmaceutical
excipients at cryogenic temperatures. Part I. Cryogenic grinding of piroxicampolyvinylpyrrolidone
mixtures. J Therm Anal Calorimetry 89:699707.
Shakhtshneider TP, Dane`de F, Capet F,
Willart JF, Descamps M, Myz SA, Boldyreva E,
97.
98.
99.
100.
101.
102.
103.
104.
105.
Boldyrev VV. 2007. Grinding of drugs with pharmaceutical excipients at cryogenic temperatures.
Part II. Cryogenic grinding of indomethacin
polyvinylpyrrolidone mixtures. J Therm Anal
Calorimetry 89:709715.
Miro A, Quaglia F, Sorrentino U, La Rotonda MI,
DEmmanuele Di Villa Bianca R, Sorrentino R.
2004. Improvement of gliquidone hypoglycaemic
effect in rats by cyclodextrin formulations. European J Pharm Sci 23:5764.
Fukami T, Furuishi T, Suzuki T, Hidaka S, Ueda
H, Tomono K. 2006. Improvement in solubility of
poorly water soluble drug by cogrinding with
highly branched cyclic dextrin. J Inclusion Phenomena Macrocyclic Chem 56:6164.
Magarotto L, Bestini S, Casentino C, Torri G.
2001. Characterization of Nimesulide/-cyclodextrin composite obtained by solid state activation.
Mater Sci Forum 360362:643648.
Grassi M, Coceani N, Magarotto L. 2002. Modelling partitioning of sparingly soluble drugs in a
two-phase liquid system. Int J Pharm 239:157
169.
Carli F, Colombo I, Torricelli C. 1987. Drug/bcyclodextrin systems by mechano-chemical activation. Chimicaoggi 5:6164.
Sugimoto M, Okagaki T, Narisawa S, Koida Y,
Nakajima K. 1998. Improvement of dissolution
characteristics and bioavailability of poorly water
soluble drugs by novel cogrinding method using
water-soluble polymer. Int J Pharm 160:1119.
Yang W, de Villiers MM. 2004. The solubilization
of the poorly water soluble drug nifedipine by
water soluble 4-sulphonic calix[n]arenes. Eur
J Pharm Biopharm 58:629636.
Sawayanagi Y, Nambu N, Nagai T. 1983. Dissolution properties and bioavailability of phenytoin
from ground mixtures with chitin or chitosan.
Chem Pharm Bull 31:20642068.
Nandi I, Bari M, Joshi H. 2003. Study of isopropyl
myristate microemulsion systems containing
cyclo-dextrins to improve the solubility of 2 model
hydrophobic drugs. AAPS PharmSciTech 4:19.
DOI 10.1002/jps

Activaton in Drugs

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Activaton in Drugs

Încărcat de

Drepturi de autor:

Formate disponibile

Drug Mechanochemical Activation

I. COLOMBO,1 G. GRASSI,2,3 M. GRASSI4

Department of Life Sciences, University of Trieste, Trieste, Italy

Received 30 September 2008; revised 12 December 2008; accepted 28 January 2009

mechanical activation; cogrinding bioavailability

mally activated chemistry can be subdivided in. In

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

COLOMBO, GRASSI, AND GRASSI

been the first type of energy deliberately employed

considerably developed and led to patents and

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

DRUG MECHANOCHEMICAL ACTIVATION

material in the form of the so-called excess Gibbs

surface. The iteration of this phenomenon induces

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

COLOMBO, GRASSI, AND GRASSI

dislocations), plane defects (e.g. grain boundaries

be equal to the surface energy, 2gsv, competing to

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

DRUG MECHANOCHEMICAL ACTIVATION

coincides with the material solidvapour surface

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

COLOMBO, GRASSI, AND GRASSI

Figure 2. When cyclodextrin is used as carrier in

of the delivery system is a prompt action, the

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

DRUG MECHANOCHEMICAL ACTIVATION

drug solubility increases for smaller crystals, see

where Tm1 is the infinite radius crystal melting

where Dhm1 is macrocrystal melting enthalpy

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

COLOMBO, GRASSI, AND GRASSI

Figure 4. Theoretical dependence of nimesulide

126.68C. It is worth mentioning that DSC can

where Dc0 is initial crystallite diameter. Another

Figure 5. Differential scanning calorimeter pattern

where k is a temperature dependent rate constant,

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

DRUG MECHANOCHEMICAL ACTIVATION

results. For these reasons synchrotron and

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

COLOMBO, GRASSI, AND GRASSI

where Cs is the time (t) dependent drug solubility,

conditions according to usual dissolution testing.

Figure 6. Amorphous drug re-crystallisation can

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

DRUG MECHANOCHEMICAL ACTIVATION

amorphous, nano-crystalline drug contents and

respect to the original Invirase1 (sanquinavir

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

COLOMBO, GRASSI, AND GRASSI

they can be subdivided into three main categories

mill lid, while the lower part is connected by

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

DRUG MECHANOCHEMICAL ACTIVATION

Attritors are made up by a vertical or horizontal

Shear Action Mills

Shock Action Mills

liners are constituted by polyurethane or dense

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

COLOMBO, GRASSI, AND GRASSI

consists of stainless steel vessel immersed in

the most important advantages of ANN consists in

where J, the number of impulses per unit time,